Lars Ackermann, M.Sc

Fulltime Academic

E-Mail:
lars.ackermann [at] uni-bayreuth.de
Telefon:
+49 (0)921 55-7623
Raum:
0.23 (Angew-Inf)
Ackermann_Lars

Lars Ackermann

 

Publication

Email Semantics – Semantic Filtering on mailboxes with respect to prevailing security policies

One frequently reaches the limits of the well-known search-by-keyword methodology. Problems are usually caused by keyword ambiguities, the lack of capabilities to model semantic relations between keywords and the absence of an underlying categorization system for relevant contents.

That is why Email Semantics addresses information extraction methods for email documents. The foundation is an ontology-based search engine which is able to localize and extract knowledge from IMAP emails. Afterwards this knowledge is used to provide smart filters in order to cope with search queries that contain complex questions. This requires, at least, answering the following questions: How can one identify sent emails where no response arrived, how can contained numeric values be categorized according to dynamic ranges and how to extract further information from attached images using the ontology-annotator principle? An important approach is the term-specific annotation in order to connect the knowledge base, i.e. an ontology, with email contents. Additionally methods for rule-based knowledge-networking are used to interconnect several email documents.

This allows for mail filtering beyond the well-established but limited keyword search. Searching for keywords yields loose text snippets – whereas semantic search yields answers.

  • ISBN-10: 3639461533
  • ISBN-13: 978-3639461534
  • Year: 2012
  • Publisher: AV Akademieverlag

Research Interests

Modeling based on Natural Language, Process Reconstruction, Information Retrieval

Publications

Lars Ackermann, Stefan Schönig, Stefan Jablonski:
Towards Simulation- and Mining-based Translation of Process Models. EOMAS, Springer Berlin Heidelberg GmbH, Ljubljana, Slovenia, 2016.

Lars Ackermann, Stefan Schönig, Stefan Jablonski:
Simulation of Multi-Perspective Declarative Process Models. BPI, Springer Berlin Heidelberg GmbH, Rio de Janeiro, Brasil, 2016.

Lars Ackermann, Stefan Schönig, Stefan Jablonski:
Towards Simulation- and Mining-based Translation of Resource-aware Process Models. REMA, Springer Berlin Heidelberg GmbH, Rio de Janeiro, Brasil, 2016.

Lars Ackermann, Stefan Schönig:
MuDePS: Multi-Perspective Declarative Process Simulation. BPM Demos, Rio de Janeiro, Brasil, 2016.

Lars Ackermann, Stefan Schönig, Michael Zeising, Stefan Jablonski:
Natural Language Generation for Declarative Process Models. 11th International Workshop on Enterprise & Organizational Modeling and Simulation (EOMAS 2015), Springer LNBIP, Stockholm, Sweden, 2015.

Lars Ackermann, Bernhard Volz:
model[NL]generation: Natural Language Model Extraction. The 13th Workshop on Domain-Specific Modeling, Indianapolis, United States, 2013.